formats and procedures to measure vocabulary knowledge. This study investigated a) the relationship between EFL learners’ self-rating of vocabulary knowledge on a newly designed self-assessment questionnaire and their performance on Vocabulary Levels Test, and b) the possibility of using self-assessment report as a valid basis for placement purposes. The results from three groups of participants (N=295) with different vocabulary proficiency levels show a) a high internal consistency among questionnaire items and b) a moderate correlation between self-assessment and VLT scores. The low group, knowing the most frequent 3,000 word families, tended to overestimate and the high group, with over 5,000 word families, tended to underestimate their vocabulary knowledge. The middle group with over 3,000 and below 4,000 word families was more realistic. Apparently, less proficient learners may have little knowledge about what they do and more proficient ones may be aware of the limits of their knowledge. This study suggests a) test developers to design innovative and systematic instruments for self-assessment, b) teachers to provide EFL learners with practice in self-assessment, and c) language testers to use self-assessment data elicitation procedures along with a valid test for grading purposes and providing complementary information.